Naturally, one would suspect that the algorithm creates images that are:
- highly unlikely to produce the same identicon twice; and
- capable of ensuring that each identicon is sufficiently distinctive as to not appear too similar to any other identicon